This pipeline is used to identify drug hits for hepatocellular carcinoma (Chen B. Gastroenterology, 2017). It can be applied to other cancers as well.
The pipeline includes four major components:
Correlate tumor samples and cell lines
Select tumor samples that are not correlated to cell lines
Select cell lines that are correlated to tumor samples
Create disease gene expression signatures
Validate signatures using external sets
Predict drugs using CMap and LINCS
Analyze drug hits including niclosamide and NEN
Examine correlation between disease gene signatures and drug gene signatures derived from in vitro/in vivo studies
Identify genes reversed by the drugs
-
dowload data from https://www.synapse.org/#!Synapse:syn6173892/files/
-
set up workspace in main.R
-
run workflow.R in each component.
Contact Bin Chen ([email protected]) for any questions.